Skip to main content

Feed

Latest Intelligence

Every AI development we have covered, newest first. Filter by section to focus on what matters to you.

SiliconANGLEBusiness

Physical AI’s moment has arrived – but moving from demo to deployment is the hard part. AWS wants to fix that

Amazon Web Services is expanding its infrastructure focus toward Physical AI to address challenges in transitioning robotics and embodied systems from experimental demos to production deployment. Physical AI applications—including vision-language-action models and world models—require processing real-time perception and physical reasoning under strict latency constraints. AWS aims to streamline the operational lifecycle, data pipelines, and edge-to-cloud infrastructure needed to support enterprise robotics and autonomous physical workflows.

41/100Intel Score, moderate impact
Ars TechnicaBusiness

Nvidia senior manager tied to ex-Supermicro staff's AI smuggling scheme

A senior manager at Nvidia has been indicted in connection with a scheme involving former Supermicro personnel to illicitly smuggle advanced AI servers and chips to China. The indictment highlights ongoing enforcement efforts by United States authorities to curb unauthorized transfers of restricted semiconductor technologies, coming after previous scrutiny and internal warnings regarding export compliance and illicit supply chain diversion.

50/100Intel Score, high impact
CNBC TechBusiness

Goldman Sachs partner warns of 'huge danger' in letting AI replace bankers' reasoning skills

A senior technology partner at Goldman Sachs has cautioned against over-relying on artificial intelligence in financial workflows, warning of significant risks if automated systems supplant the foundational reasoning skills of junior bankers. While the investment bank continues adopting generative AI tools to boost operational efficiency, the leadership warning highlights growing institutional concern that delegating critical analytical tasks to algorithms could degrade core problem-solving capabilities, domain expertise, and professional talent development across the financial services sector over time.

31/100Intel Score, moderate impact
SiliconANGLEBusiness

Thomson Reuters launches proprietary AI model for legal work

Thomson Reuters has introduced Thomson, its first proprietary large language model tailored for legal workflows. The domain-specific model combines the company's curated legal content assets with third-party foundation models. Thomson will initially power the Tabular Analysis feature within the CoCounsel Legal assistant, designed for high-volume document review and analysis tasks. The release marks an evolution in Thomson Reuters' strategy from relying solely on external model providers to deploying proprietary vertical models.

43/100Intel Score, moderate impact
Financial Times (AI)Business

Can AI glasses replace the smartphone?

Financial Times reports on Big Tech's push into AI-powered smart glasses and wearable devices as potential successors or primary interfaces for smartphones, amidst persistent public pushback and privacy concerns labelling the technology as intrusive stalkerware.

42/100Intel Score, moderate impact
SiliconANGLEBusiness

Sam Altman voices fears that control of AI could be centered in too few hands

In an interview with podcaster David Senra, OpenAI Chief Executive Sam Altman expressed concern over the potential centralization of artificial intelligence development and governance. Altman warned that if control over leading AI technologies becomes concentrated among a small number of companies or individuals, broader society will lose influence over how the technology evolves and affects economic structures. The remarks touch on ongoing debates regarding compute concentration, market competition, and the role of open-weight models.

32/100Intel Score, moderate impact
SiliconANGLEBusiness

Report: AI model hub Hugging Face exploring sale at $13B valuation

Hugging Face is reportedly exploring a potential sale that could value the artificial intelligence repository and developer platform at $13 billion or more, according to a report from Business Insider. The company has reportedly engaged an investment bank to gauge interest from potential buyers. Deliberations remain in early stages, and no formal bids or prospective acquirers have been publicly confirmed. Hugging Face functions as a foundational registry and collaboration hub for open-source AI models, datasets, and machine learning development workflows.

61/100Intel Score, high impact
TechCrunchBusiness

Is it legal to train AI models on copyrighted books? It’s complicated

Legal and ethical questions surrounding the use of copyrighted literature to train generative AI models remain unresolved. AI developers have routinely used large corpora of scraped and digitized books without explicit licensing or author consent. While creators argue this practice infringes on intellectual property rights and disrupts publishing livelihoods, technology companies often cite fair use doctrines to defend training practices across the AI industry.

34/100Intel Score, moderate impact
TechCrunchBusiness

Harvard’s $699 startup bootcamp offers AI avatars of its instructors

Harvard Business School has introduced AI-driven instructor avatars within its $699 HBS Foundry startup bootcamp. The synthetic avatars are deployed to simulate interactive scenarios, delivering automated feedback to participants during practice pitch sessions and simulated board meetings. The program represents an institutional deployment of generative persona modeling to scale interactive professional coaching at a lower price point than traditional executive education courses.

20/100Intel Score, low impact
SiliconANGLEBusiness

From tokenmaxxing to sovereign alpha: Who controls your AI economics?

An analysis of enterprise artificial intelligence economics argues that vendor-centric metrics such as token volume and API calls fail to reflect business value. Highlighting reporting that Canva revised its 2026 revenue-growth forecast downward from 30% to 20% due to surging AI operating expenses, the piece emphasizes the need for enterprises to gain control over inference costs, unit economics, and architectural sovereignty rather than pursuing unconstrained model usage.

44/100Intel Score, moderate impact